Skip to main content
Meo

Most cost-efficient AI models ($ per correct answer)

Ranked by USD per correct answer — true intelligence-per-dollar, not just sticker price. A cheap model that needs many retries is not cheap. Cost-per-correct already amortizes the error rate, so it is the honest cost metric for budget-sensitive workloads.

Ranking 82 models across the meo-tested roster · as of 2026-10-18.

#ModelLab$/correct
1Owl Alphaopenrouter$0.0000
2Nex AGI: Nex-N2-Pro (free)nex-agi$0.0000
3NVIDIA: Nemotron 3 Nano Omni (free)nvidia$0.0000
4NVIDIA: Nemotron Nano 12B 2 VL (free)nvidia$0.0000
5Google: Gemma 3 4Bgoogle$0.0001
6Google: Gemma 3 12Bgoogle$0.0002
7Amazon: Nova Lite 1.0amazon$0.0002
8Google: Gemma 3 27Bgoogle$0.0003
9Mistral: Ministral 3 3B 2512mistralai$0.0004
10Meta: Llama 4 Scoutmeta-llama$0.0005
11Mistral: Ministral 3 8B 2512mistralai$0.0005
12Mistral: Mistral Small 3.2 24Bmistralai$0.0006
13Mistral: Ministral 3 14B 2512mistralai$0.0007
14OpenAI: GPT-4.1 Nanoopenai$0.0008
15Anthropic: Claude 3 Haikuanthropic$0.0009
16Perceptron: Perceptron Mk1perceptron$0.0009
17Qwen: Qwen2.5 VL 72B Instructqwen$0.0009
18Meta: Llama 4 Maverickmeta-llama$0.0010
19Qwen: Qwen3 VL 32B Instructqwen$0.0011
20ByteDance Seed: Seed-2.0-Minibytedance-seed$0.0013
21MiniMax: MiniMax-01minimax$0.0014
22Xiaomi: MiMo-V2.5xiaomi$0.0016
23DeepSeek: DeepSeek V4 Flash 0423deepseek$0.0016
24OpenAI: GPT-4o-miniopenai$0.0019
25ByteDance Seed: Seed 1.6 Flashbytedance-seed$0.0022
26Qwen: Qwen3 VL 8B Instructqwen$0.0022
27Qwen: Qwen3 VL 30B A3B Instructqwen$0.0023
28OpenAI: GPT-4.1 Miniopenai$0.0023
29Meta: Llama 3.2 11B Vision Instructmeta-llama$0.0023
30inclusionAI: Ring-2.6-1Tinclusionai$0.0024
31Mistral: Mistral Large 3 2512mistralai$0.0024
32Qwen: Qwen3.5-9Bqwen$0.0026
33Qwen: Qwen3 VL 235B A22B Instructqwen$0.0028
34OpenAI: GPT-5.4 Nanoopenai$0.0029
35Google: Gemma 4 31Bgoogle$0.0030
36Z.ai: GLM 4.6Vz-ai$0.0035
37OpenAI: GPT-6.1 Solopenai$0.0036
38Google: Gemini 2.5 Flash Litegoogle$0.0037
39Google: Gemma 4 26B A4B google$0.0043
40Mistral: Mistral Small 4mistralai$0.0046
41MiniMax: MiniMax M3minimax$0.0047
42OpenAI: GPT-5 Nanoopenai$0.0047
43Tencent: Hy3 previewtencent$0.0051
44Qwen: Qwen3.5-Flashqwen$0.0051
45Qwen: Qwen3 VL 30B A3B Thinkingqwen$0.0068
46ByteDance Seed: Seed 1.6bytedance-seed$0.0069
47ByteDance Seed: Seed-2.0-Litebytedance-seed$0.0080
48OpenAI: GPT-5.1-Codex-Miniopenai$0.0082
49Qwen: Qwen3.7 Plusqwen$0.0083
50Baidu: ERNIE 4.5 VL 424B A47B baidu$0.0087
51OpenAI: GPT-5 Miniopenai$0.0102
52Arcee AI: Trinity Large Thinkingarcee-ai$0.0108
53Perplexity: Sonarperplexity$0.0119
54SpaceXAI: Grok 4.3x-ai$0.0127
55Google: Gemini 3.1 Flash Lite Previewgoogle$0.0128
56Qwen: Qwen3 VL 8B Thinkingqwen$0.0138
57Qwen: Qwen3.6 Plusqwen$0.0138
58MoonshotAI: Kimi K2.7 Codemoonshotai$0.0139
59Qwen: Qwen3.5 Plus 2026-02-15qwen$0.0147
60Qwen: Qwen3.7 Maxqwen$0.0157
61Qwen: Qwen3.6 Flashqwen$0.0157
62StepFun: Step 3.7 Flashstepfun$0.0162
63Anthropic: Claude Opus 5.5anthropic$0.0165
64DeepSeek: DeepSeek V4 Pro 0423deepseek$0.0182
65Qwen: Qwen3.6 35B A3Bqwen$0.0189
66Google: Gemini 3.1 Flash Litegoogle$0.0191
67Z.ai: GLM 5.2z-ai$0.0200
68NVIDIA: Nemotron 3 Ultranvidia$0.0224
69Qwen: Qwen3.5-35B-A3Bqwen$0.0228
70Xiaomi: MiMo-V2.5-Proxiaomi$0.0244
71Google: Gemini 4 Argongoogle$0.0246
72MoonshotAI: Kimi K2.6moonshotai$0.0281
73OpenAI: GPT-5.5openai$0.0336
74Qwen: Qwen3.5-27Bqwen$0.0348
75MoonshotAI: Kimi K3moonshotai$0.0379
76Z.ai: GLM 5.1z-ai$0.0426
77Google: Gemini 3.5 Flashgoogle$0.0432
78Anthropic: Claude Opus 4.8anthropic$0.0459
79Anthropic: Claude Fable 5anthropic$0.0573
80SpaceXAI: Grok 4.7x-ai$0.0579
81Google: Gemini 3.1 Pro Previewgoogle$0.0597
82OpenAI: GPT-5.5 Proopenai$0.3197

← All rankingsMethodology & 𝕍 →